# Gasahol Equation

Discussion in 'Electronic Design' started by Jim Weir, Jul 17, 2004.

1. ### Jim WeirGuest

I'm making a % alcohol meter for gasahol. It is an article for my monthly
column in Kitplanes magazine.

However, it has been a very long time since I've done a ratiometric equation, so
would somebody please check my work? Many thanks.

I start with an air variable capacitor of 365 pf (the old crystal set tuner).

I dunk it in pure gasoline with a dielectric constant of 1.94 and I should come
up with a 708 pf capacitor.

I then mix 90 ml of pure gasoline (e = 1.94) with 10 ml of pure ethyl alcohol (e
= 24.5) and come up with a composite dielectric constant of 4.2, which gives my
capacitor dunked in this mixture a value of 1533 pf ... (which then lowers the
audio frequency of the oscillator that the capacitor is controlling, which then
goes to the frequency to voltage converter, which goes out to the Rat Shack
panel meter calibrated in percentage...)

Did I do that right? If that is correct, then all the rest of the percentages I
calculated on the spreadsheet from 1 to 10 should be OK also.

Jim

2. ### John WoodgateGuest

Looks OK to me.

3. ### Tim WescottGuest

"Did I do that right" and "metrology" never seem to go together well. I
think the more you know the less you're willing to say it must be right.

* Are you sure that the relationship is linear? It sounds like you've
tested it for a couple of mixtures which is good, it may be a good idea
to test a bit more (perhaps a 50% mixture).

* What happens if you measure wet gasahol, i.e. gasahol that's sat
around for a bit with that ever so hydroscopic alcahol in it -- will it
get conductive and change the frequency significantly? Do you have an
independent way of checking? Your capacitor will certainly have a great
deal of surface area for a conductive liquid to contact.

4. ### ScrimGuest

I was asked once to build a device for measuring the dielectric constant of
ink. The ink was an oil/water emulsion and I never really fully understood
the consequences of the fact this medium was conductive. To try and reduce
its effect I used a capacitor in series with the test cell, but whilst that
removed any dc bias, it wasn't a full solution. The laboratory that used the
device told me some time later that it turned out to be very good at
measuring some other property of emulsions, even if it didn't really measure
dielectric constant. I'm not sure if alcohol is conductive, but unless
specially prepared as anhydrous, even 'pure' alcohol has a couple of percent
of water in it and its conductivity may vary greatly with concentration of
dissolved impurities.

Scrim

5. ### Don LancasterGuest

Um, I believe "pure" alcohol is 180 proof, or ten percent water.

--
Many thanks,

Don Lancaster
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
voice: (928)428-4073 email:

Please visit my GURU's LAIR web site at http://www.tinaja.com

6. ### ScrimGuest

Well, ok, to be precise, the strongest ethanol/water mix that can be
produced by simple distillation is 95.6% ethanol, 4.4% water. This is an
azeotropic mixture, which means that when distilled, the vapour contains the
constituents in the same proportion as the liquid and so can't result in any
greeater concentration.

Scrim

7. ### Robert BaerGuest

By definition, pure alcohol is 200 proof, and one laboratories can get
white tax stamp (untaxed) Triple Distilled Ethanol which (if i remember
correctly) has under 1% water (the main "contaminant").

8. ### Rich GriseGuest

Everybody seems to be concerned about the water being conductive.
Water isn't a conductor. [0]
alcohol isn't a conductor.
gasoline isn't a conductor.

What part of "isn't a conductor" is the conductivity coming from?

As far as the calculations, I'd have to see more data points to
get a better idea of the transfer function, but even with them, I'm
way too lazy to do that much arithmetic. I'd calibrate it by doing
measurements on known mixes, as you seem to be doing. Then,
who needs math when 1048576-byte EPROMS are a few bucks? Can you say
look-up table?
--
Cheers!
Rich

[0] it's the dissolved ions. And there probably aren't a lot
of ion vapor molecules floating around in the air to get sucked
up by the ethanol.

9. ### Boris MoharGuest

Post that to sci.physics and watch for a reply from Uncle Al.

10. ### JimmieGuest

The circuit he is using could be used to determine the amount of H2O present
in the alcohol. The main problem with the water is not that it may be
conductive but that it has a very high dielectric constant due to it being a
polar molecule that can invalidate the readings.

11. ### ScrimGuest

Why is its high dielectric constant a problem?

Scrim

12. ### Tim WescottGuest

Because he's deducing the alcohol to gasoline ratio by the dielectric
constant of the mixture. Put a 3rd ingredient in there and you have one
equation and two unknowns -- so you don't know your ratio any more.

13. ### JimmieGuest

Dielectric constant (DC)of h20 is about 80, gasoline about 2 and his mix of
alcohol about 25. Doesnt take much extra water to really skew the results.
This may not be problem though as I suspect the alcohol the OP is using
normally has a significant amount of water already in it per the quoted DC.
a little more may not make a lot of difference.